Ridley: Consensus is another word for groupthink

Blood-letting. Freud. Climate science…

If I give you a series of numbers “2, 4, 6” and ask what number comes next you will probably reply “8”. I say yup—and so what is the rule I am following? You will probably reply “add 2” or “ascending even numbers”. Nope, I say, wrong. My rule was simpler: “ascending numbers”. So “7” and “9” and “100” would have also been fine.

This was the example given by the psychologist Peter Wason in 1960 when he coined the term “confirmation bias”. What you should have done to test your hypothesis that my rule was “ascending even numbers” was to offer an odd number, not an even number. It is more useful to try to falsify your own hypothesis rather than confirm it. Very few of us do this.

Wason later did an even more striking experiment. He presented subjects with four cards, each of which had a letter on one side and a number on the other: The visible side of the four cards read D, F, 3, 7. Which two cards do you have to turn over to check the rule “if a card has a D on one side, it has a 3 on the other”? Most people get this wrong. The answer is at the end of this article*.

Confirmation bias is the mother of all psychological flaws in human reasoning. It infects everybody, myself included. It is rampant in economics and ubiquitous in science. It is not a sin or a crime, but it is a frequent and widespread cause of people being misled—and therefore of bad policy. “Of the fifty-odd biases discovered by [Daniel] Kahneman, [Amos] Tversky, and their successors, forty-nine are cute quirks, and one is destroying civilization,” writes Scott Alexander, founder of the rationalist blog Slate Star Codex. “This last one is confirmation bias - our tendency to interpret evidence as confirming our pre-existing beliefs instead of changing our minds.”

Take the performance of the British economy since Brexit. Leavers see a decade of sluggish growth as evidence that forecasts of a post-Brexit collapse were wrong. Remainers see a decade of sluggish growth as evidence that Brexit harmed the economy.

“Looking at how well UK trade and GDP has held up relative to our big EU rivals, we can be confident any negative effect has been small at worst,” writes David Paton, professor of industrial economics at Nottingham University

“One apparent puzzle,” writes the economist Jonathan Portes, “is that the aggregate data do not always show a simple collapse in UK-EU trade relative to trade with the rest of the world. But this is less reassuring than it looks [because] aggregate trade flows can look relatively resilient while the number of firms exporting to the EU falls”.

When any statistic can be recruited to support your preferred theory, you are suffering from confirmation bias. Scientists like to boast that they strive to prove themselves wrong, that following the philosopher Karl Popper they know falsification is the essence of science. Nonsense. Like prosecutors determined to find a suspect guilty, they usually seek out evidence that bolsters, rather than challenges, their pet ideas. In this they are only human.

As the philosopher Francis Bacon noted in 1620: “And such is the way of all superstition, whether in astrology, dreams, omens, divine judgments, or the like; wherein men, having a delight in such vanities, mark the events where they are fulfilled, but where they fail, though this happen much oftener, neglect and pass them by.”

So how then does science work at all? Not by scientists trying to falsify their own theories but by them trying to falsify each other’s theories. A vital part of science is that it has been decentralised, both geographically and financially, so that Prof B has no hesitation in arguing that Prof A is talking through his test tube.

This has now changed. Funding flows through a few highly centralised pipes; and in field after field there is an imposed “consensus” as to which hypothesis is right. Scientists, like lawyers and doctors, become more confident in their judgment as they become more senior, requiring less positive evidence to stick to their opinions than they need negative evidence to abandon them. When a discipline defers to authority, and demands adherence to a creed, it becomes a cult. Medicine did this with blood-letting; psychoanalysis did it with Freud; climate science with ocean acidification; Alzheimer’s research with amyloid plaques. Consensus is another word for groupthink.

An especially acute example of confirmation bias is the widespread assumption in social science that parenting style determines the personality of offspring. People assert, as if it were an established fact, that somebody is aggressive because their father was aggressive to them in childhood. It is so obvious that it does not need to be proved.

But it turns out this notion is as dodgy as saying “8” proves the rule of ascending even numbers. Yes, aggressive fathers have aggressive children but how do we know whether this is causal rather than because they share many genes? Accumulating examples of correlation will not help. We need to find cases where children of aggressive fathers are raised by non-biological parents and where non-biological children are raised by aggressive fathers.

In her seminal 1998 book The Nurture Assumption, the psychologist Judith Rich Harris chronicled her own eye-opening discovery that the overwhelming evidence from identical twins raised apart and adopted children raised together comprehensively falsifies the theory that (outside cases of extreme abuse) parenting style has lasting effects on children’s personality.

It is not all genes. Children get their personality partly from their peers as well as from their genes—but hardly at all from the way their parents treat them. This is true not just of aggression but many other features of personality. Yet in the three decades since Harris’s book came out, has social science, or popular media, moved away from its groupthink on the matter? Hardly at all.

*You should turn over D and 7.